Class 1a. Introduction to the enterprise

1
CAS LX 522Syntax I

• Class 1a.Introduction to the enterprise

2
Some things we know

• Is this English?
• The cat slept.
• Slept the cat.
• Cat slept the.
• Cat the slept.
• Why?

3
The task

• What do we know?
• The comes before cat, cat comes before slept.
• Try to generalize.
• Slept is the verb, maybe this holds of all verbs.
• The cat is the subject, maybe this holds of all
  subjects.
• Subjects contain the and a noun, with the first.
• An English sentence has a subject followed by a
  verb.
• Formalize (make precise)
• Nouns cat, dog
• Verbs slept, yawned
• Sentence Subject the Noun Verb

4
The task

• Check
• Sentence Subject the Noun Verb
• The cat slept.
• The dog yawned.
• The cat yawned.
• The dog slept.
• Look at further data (predictions)
• The cat chased the dog.
• This is an English sentence, but our schema
  cannot produce it. Our theory of English
  sentences is insufficient. We need to
  revise/extend it.

5
The task

• Consider the counterexample (or the class of
  counterexamples) to understand where the current
  theory falls short.
• The cat chased the dog.
• The dog is probably the same kind of thing as the
  cat, but we dont want to call it a subject
  (its traditionally called the object).
• It contains the and a noun, and the noun seems to
  be the most important part.
• Since it contains more than one word, we can call
  it a phraseits not a whole sentence, but its
  more than a word.
• So, well call it a noun phrase.

6
The task

• Consider the counterexample (or the class of
  counterexamples) to understand where the current
  theory falls short.
• The cat chased the dog.
• In this English sentence, there is a noun phrase
  both before and after the verb. So, in addition
  to our previous schema, we add a second one.
• Theory of English sentences
• Sentence NP the Noun Verb
• Sentence NP the Noun Verb NP the Noun

7
Lather, rinse, repeat

• And the process continues.
• The cat chased a dog.
• A cat chased the dog.
• A cat chased a dog.
• It looks like a NP can either have the or a as
  its first element. Thus
• Theory of English sentences
• Sentence NP the Noun Verb
• Sentence NP a Noun Verb
• Sentence NP the Noun Verb NP the Noun
• Sentence NP the Noun Verb NP a Noun
• Sentence NP a Noun Verb NP the Noun
• Sentence NP the Noun Verb NP a Noun

8
Generalizing

• What weve ended up with is a bit clumsy, but we
  can now generalize our schemas to make this more
  compact
• NP the Noun
• NP a Noun
• Sentence NP Verb
• Sentence NP Verb NP
• Not only does this reduce the amount we have to
  write down, but it actually makes a more profound
  prediction If this much of our theory of English
  sentences is right, then anything that can be a
  noun phrase subject can also be a noun phrase
  object. This is not just making our notation more
  compact, but it is a substantive addition to the
  theory.

9
Compacting the notation

• There are some further ways we can consolidate
  our theory of English sentences by using some
  common notational tools.
• X is optional (X)
• Either Y or Z Y/Z
• Thus
• Sentence NP Verb (NP)
• NP the/a Noun
• Unlike our introduction of a separate schema for
  NP, this change is not a substantive change to
  our theory of English sentences, it is just a
  shorthand for the same theory.

10
The grumpy cat

• As a demonstration of the benefit of introducing
  a separate NP schema, consider
• The grumpy cat chased the unhappy dog.
• How can we extend our theory of English sentences
  to allow for this sentence? What other word
  sequences are predicted to be English sentences?
  Are they?

11
Now, what are we doing?

• Ok, so we have the beginnings of a theory of
  English sentences. But what is it?
• As weve developed it, it is a description of
  sentences of English, what we might need if we
  wanted to program a computer to produce English
  sentences.
• But it is also a subset of what English speakers
  know about English.
• You may or may not have previously thought about
  the fact that subjects precede verbs and objects
  follow verbs (or the analog in your native
  language), but you knew it nevertheless. You
  could identify sequences of words that did not
  have this property as not being part of your
  language, but its tacit knowledge. As such, we
  have to study this knowledge indirectly, based on
  what are judged to be valid sentences and what
  arent.

12
What English speakersknow about English

• An English speaker has a complex system of
  knowledge that allows him/her to distinguish
  between sentences of English and non-sentences of
  English. Well refer to this system as a grammar.
  At its simplest, a grammar is a means of deciding
  whether a sequence of words is grammatical (e.g.,
  a sentence of English) or not. Were studying the
  properties of that system.
• Its not always obvious what it is that is wrong
  with non-sentences, but still the judgments
  (intuitions) are clear.

13
Types of (un)acceptability

• Big that under staple run the jump swim.
• The dog are snoring.
• These are ungrammaticalthere is a problem with
  their form, they are not English. We write to
  indicate this.
• My toothbrush is pregnant again.
• This is nonsensical, given our knowledge about
  the world (not about English), but it is
  grammatical.
• As I knitted the sock The horse raced fell to
  the floor. past the barn fell.
• The rat the cat the dog chased caught escaped
  adeptly.
• These are interestingly difficult to parse but
  once you get it, they are fine (if clumsy)
  sentences of English.

14
Parentheses and optionality

• In describing data, people will often use the (),
  shorthand notation to indicate optionality or
  options
• Pat (quickly) ran to the bank.
• Pat ran to the bank. Pat quickly ran to the
  bank.
• Pat washed (quickly) the asparagus.
• Pat washed the asparagus. Pat washed quickly the
  asparagus.
• The dish ran away with (the) spoon.
• The dish ran away with the spoon.
• The dish ran away with spoon.
• The cat chased a/the dog.
• The cat chased a dog. The cat chased the dog.

15
Ambiguity and stars

• Sentences can be ambiguous.
• I sat by the bank.
• Sometimes we might have reason to expect
  ambiguity that is not there, which is also
  indicated using , on a disambiguating
  continuation.
• How did John say Mary fixed the car?
• With a wrench.
• In a high-pitched voice.
• How did John ask if Mary fixed the car?
• With a wrench.
• In a high-pitched voice.

16
Knowledge of language is actually really
complicated

• Bill told her mother that Mary is a genius.
• Bill told her that Mary is a genius.
• I told Mary that Pat gave a book to me.
• Who did I tell that Pat gave a book to me?
• Who did I tell Mary that gave a book to me?
• Who did I tell Mary that Pat gave a book to?
• I loaned Mary the book Pat gave me.
• Who did I loan the book Pat gave me?
• Who did I loan Mary the book gave me?
• Who did I loan Mary the book Pat gave?

17
How do people know this?

• All native speakers of English know this.
• Little kids werent told these rules (or punished
  for violating them)
• You cant question a subject in a complement
  embedded with that
• You cant use a proper name as an object if the
  subject is co-referential.

18
Two questions

• What do people know about their language?
• Including things we know unconsciously
• How do people come to know it?
• Tricky question for things that we dont know we
  know.

19
Systematicity

• What people eventually end up with is a system
  with which they can produce (and rate) sentences.
  A grammar.
• Even if youve never heard these before, you know
  which one is English and which one isnt
• Eight very lazy elephants drank brandy.
• Eight elephants very lazy brandy drank,
• Kids say wugs.

20
Positive and negative evidence

• Adults know if a given sentence S is grammatical
  or ungrammatical. This is part of the knowledge
  kids gain through language acquisition.
• Kids hear grammatical sentences(positive
  evidence)
• Kids are not generally told which sentences are
  ungrammatical(no negative evidence)

21
Positive and negative evidence

• One of the striking things about child language
  is how few errors they actually make.
• For negative feedback to work, the kids have to
  make the errors (so that it can get the negative
  response).
• But they dont make the errors.
• (Kids do make errors, but not of the kind that
  one might expect if they were just trying to
  extract patterns from the language data they hear)

22
Poverty of the stimulus

• What is the next number in this sequence?
• 1, 2, 3, __
• How do you form a yes-no question?
• Pat will leave.
• Will Pat leave?
• The book that you were reading was good.
• Book the that you were reading was good?
• Were the book that you reading was good?
• Was the book that you were reading good?

23
The Language instinct

• The linguistic capacity is part of being human.
• Like having two arms, ten fingers, a vision
  system, humans have a language faculty.
• The language faculty (tightly) constrains what
  kinds of languages a child can learn.
• Universal Grammar (UG).

24
But languages differ

• English, French Subject Verb Object (SVO)
• John ate an apple.
• Pierre a mangé une pomme.
• Japanese, Korean Subject Object Verb (SOV)
• Taroo-wa ringo-o tabeta.
• Chelswu-ka sakwa-lul mekessta.
• Irish, Arabic (VSO), Malagasy (VOS),

25
But languages differ

• English Adverbs before verbs
• Mary quickly eats an apple.
• (also Mary ate an apple quickly)
• Mary eats quickly an apple.
• French Adverbs after verbs
• Geneviève mange rapidement une pomme.
• Geneviève rapidement mange une pomme.

26
Parameters

• We can categorize languages in terms of their
  word order SVO, SOV, VSO.
• This is a parameter by which languages differ.
• The dominant formal theory of first language
  acquisition holds that children have access to a
  set of parameters by which languages can differ
  acquisition is the process of setting those
  parameters.
• What are the parameters?
• What are the universal principles of grammar?

27
The enterprise

• The data we will primarily be concerned with are
  native speaker intuitions.
• Native speakers, faced with a sentence S, know
  whether the sentence S is part of their language
  or isnt. These intuitions are highly systematic.
• We want to uncover the system (which is
  unconscious knowledge) behind the intuitions of
  native speakerstheir knowledge of language.

28
I-language

• We are studying the system behind one persons
  pattern of intuitions.
• Speakers growing up in the same community have
  very similar knowledge, but language is an
  individual thing (I-language).
• One doesnt need to ask the académie française
  whether Geneviève rapidement mange une pomme is a
  sentence of French. One knows.
• I-languages of a community is can be
  characterized, but it is external to the speaker
  (E-language), not any one persons knowledge, a
  generalization over many peoples I-languages.
• For example, Parisian French.

29
Competence

• We are also concerned with what a person
  knowswhat characterizes a persons language
  competence. We are in general not concerned here
  with how a person ends up using this knowledge
  (performance).
• You still have your language competence when you
  are sleeping, in the absence of any performance.
  Being drunk doesnt make one think bought some
  John coffee is English, though perhaps one might
  say it.

30
Prescriptive rules

• Another thing we need to be cautious of are
  prescriptive rules. Often prescriptive rules of
  good grammar turn out to be impositions on our
  native grammar which run counter to our native
  competence.
• After all, why did they need to be rules in the
  first place?

31
Prescriptive rules

• Prepositions are things you dont end a sentence
  with.
• It is important to religiously avoid splitting
  infinitives.
• Remember Capitalize the first word after a
  colon.
• Dont be so immodest as to say I and John left
  say John and I left instead.
• Impact is not a verb.
• The book which you just bought is offensive.

32
Prescriptive rules

• When making grammaticality judgments (or when
  asking others to make grammaticality judgments),
  we must do our best to factor out prescriptive
  rules (learned explicitly, e.g., in school).
• Were not interested in studying the prescriptive
  rules we could just look them up, and it isnt
  likely to tell us anything deep about the makeup
  of the human mind. Theyre really just a secret
  handshake, allowing educated people to detect
  one another.

33
Syntax as science

• Syntax, as practiced here, is a scientific
  enterprise. This means, in particular,
  approaching syntax using the scientific method.
• Step 1 Gather observations (data)
• Step 2 Make generalizations
• Step 3 Form hypotheses
• Step 4 Test predictions made by these
  hypotheses, returning to step 1.

34
Syntax as science

• This is pretty much the way other scientific
  disciplines work biology, chemistry, physics.
• We may start out with a kind of folk
  understanding of a field.
• For example, you push something and it moves. You
  stop pushing, and it stops. The sun revolves
  around the earth from East to West, followed by
  the moon. Water is a basic element, like fire.
  Whales are very big fish, like dolphins, or tuna,
  but bigger.
• Ockhams Razor posit as few concepts and
  relations as we can get away with. A leaner
  theory is a better theory. A more easily
  falsifiable theory is a better theory too.

35
Levels of adequacy

• If our hypotheses can predict the existence of
  the grammatical sentences in a corpus (a set of
  grammatical sentences), it is observationally
  adequate.
• Note the grammar described by some number of
  words appear in some order is observationally
  adequate, for pretty much any language. This is
  not a very difficult or satisfying level of
  adequacy to reach. Nor is it disprovable, but it
  hasnt really advanced our understanding of the
  world.
• If our hypotheses can predict the native-speaker
  intuitions about which sentences are grammatical
  and which are ungrammatical, it is descriptively
  adequate.

36
Levels of adequacy

• If we can take a descriptively adequate set of
  hypotheses one step further and account not only
  for the native speaker judgments but also for how
  children come to have these judgments, our
  hypotheses are explanatorily adequate.
• Its this last level that we are hoping to
  achieve.
• Basic principles
• Parameters of variation
• How to set the parameters from childs input

37
Infinite use of finite means

• English has an infinite number of sentences. Any
  natural language does.
• John said that English has an infinite number of
  sentences.
• Mary said that John said that English has an
  infinite number of sentences.
• Pat said that Mary said that John said that
  English has an infinite number of sentences.
• Tracy said that Pat said that Mary said that John
  said that English has an infinite number of
  sentences.
• Chris said that Tracy said that Pat said that
  Mary said that John said that English has an
  infinite number of sentences.
• If S is a sentence and N is a name,N said that S
  is also a sentence.
• S ? N said that S
• Some of the earliest work in grammatical theory
  was done by trying to state rules of this form,
  the goal being to generate the sentences of a
  language.

38
Of the past and the future

• Serious scientific study of sentence structure of
  this kind generally began in the 50s, driven to
  a large extent by the work of Noam Chomsky.
• Its now half a century later, and we have
  learned a lot about how syntax works.

39
Of the past and the future

• Progress was incremental, and often required
  revising our assumptions about how sentences are
  really put together.
• Data was examined, generalizations were arrived
  at, hypotheses were formed, predictions were
  testedand often led to revisions of the
  generalizations and the hypotheses, and so forth.

40
Of the past and the future

• Two goals of the class
• Think like a syntactician.
• Be able to read (relatively recent) books,
  articles, etc. about syntax.
• Its not really enough to just know what people
  concluded, we need to understand why they
  concluded what they did.

41
Some milestones

• Until about the mid-70s, phrase structure rules.
• S ? NP VP VP ? V (NP)
• Mid-70s, X-Bar Theory (a generalization about
  what are possible PSRs).
• In the 80s, a fairly significant shift to
  Government and Binding Theory (viewing grammar a
  little less like a computer program). Very
  productive.
• In the 90s, another shift to the Minimalist
  Program (an attempt at simplification, as well as
  a change in philosophy).

42
?

• ? ?
• ?
• ? ?
• ? ?
• ?
• ?